| 1 |
On 07/08/2018 02:50 PM, Aaron W. Swenson wrote: |
| 2 |
> On July 8, 2018 5:38:48 PM EDT, Zac Medico <zmedico@g.o> wrote: |
| 3 |
> |
| 4 |
> On 07/08/2018 02:18 PM, Michał Górny wrote: |
| 5 |
> |
| 6 |
> W dniu nie, 08.07.2018 o godzinie 14∶11 -0700, użytkownik Zac Medico |
| 7 |
> napisał: |
| 8 |
> |
| 9 |
> On 07/08/2018 01:18 PM, Zac Medico wrote: |
| 10 |
> |
| 11 |
> On 07/08/2018 01:08 PM, Michał Górny wrote: |
| 12 |
> |
| 13 |
> W dniu nie, 08.07.2018 o godzinie 11∶57 -0700, |
| 14 |
> użytkownik Zac Medico |
| 15 |
> napisał: |
| 16 |
> |
| 17 |
> On 07/08/2018 11:42 AM, Michał Górny wrote: |
| 18 |
> |
| 19 |
> W dniu nie, 08.07.2018 o godzinie 11∶04 |
| 20 |
> -0700, użytkownik Zac Medico |
| 21 |
> napisał: |
| 22 |
> |
| 23 |
> On 07/08/2018 06:56 AM, Michał Górny wrote: |
| 24 |
> |
| 25 |
> W dniu nie, 08.07.2018 o godzinie |
| 26 |
> 15∶02 +0200, użytkownik Kristian |
| 27 |
> Fiskerstrand napisał: |
| 28 |
> |
| 29 |
> On 07/08/2018 08:53 AM, Michał |
| 30 |
> Górny wrote: |
| 31 |
> |
| 32 |
> Is safe git syncing |
| 33 |
> implemented already? If not, |
| 34 |
> maybe finish it first and |
| 35 |
> cover both with a single |
| 36 |
> news item. Git is going to |
| 37 |
> be more efficient here, so |
| 38 |
> people may want to learn |
| 39 |
> they have an alternative. |
| 40 |
> |
| 41 |
> |
| 42 |
> Why complicate things, and |
| 43 |
> increase wait for something that |
| 44 |
> benefits |
| 45 |
> most users, just to give |
| 46 |
> alternatives to a few using |
| 47 |
> non-default sync |
| 48 |
> mechanism. Securing git |
| 49 |
> distribution is a whole |
| 50 |
> different ballpark. |
| 51 |
> |
| 52 |
> |
| 53 |
> |
| 54 |
> Let me rephrase. Let's say I'm using |
| 55 |
> rsync. This new feature is |
| 56 |
> something positive but it breaks my |
| 57 |
> use case (for one of the listed |
| 58 |
> reasons -- overlayfs, inode use, |
| 59 |
> small fs cache). After reading this |
| 60 |
> news item, I learn that my only |
| 61 |
> option is to disable the new feature. |
| 62 |
> |
| 63 |
> Now, I would appreciate being told |
| 64 |
> that there's an alternate sync method |
| 65 |
> that handles secure updates without |
| 66 |
> having all those drawbacks. |
| 67 |
> |
| 68 |
> |
| 69 |
> The thing is, the normal git tree |
| 70 |
> doesn't even provide pre-generated |
| 71 |
> metadata, and I see then gentoo-mirror |
| 72 |
> repo that provides metadata does |
| 73 |
> not have commits signed with an release key: |
| 74 |
> |
| 75 |
> https://github.com/gentoo-mirror/gentoo/commits/stable |
| 76 |
> |
| 77 |
> So I'm really not comfortable |
| 78 |
> recommending git to anyone at this point. |
| 79 |
> |
| 80 |
> |
| 81 |
> Wrong twice. |
| 82 |
> |
| 83 |
> Firstly, the canonical URL is: |
| 84 |
> |
| 85 |
> https://anongit.gentoo.org/git/repo/sync/gentoo.git |
| 86 |
> (https://gitweb.gentoo.org/repo/sync/gentoo.git) |
| 87 |
> |
| 88 |
> Secondly, the merge commits (i.e. top |
| 89 |
> commits that are verified |
| 90 |
> by Portage) are signed by dedicated key that |
| 91 |
> is part of the infra key |
| 92 |
> set. In other words, it works out of the box. |
| 93 |
> |
| 94 |
> |
| 95 |
> Is there any documentation that shows users how |
| 96 |
> to migrate to git, and |
| 97 |
> what the pros and cons might be? Maybe its |
| 98 |
> worthy of its own news item. |
| 99 |
> |
| 100 |
> |
| 101 |
> Maybe. I don't really know, and don't think it's a |
| 102 |
> good idea to show 30 |
| 103 |
> news item of things users might like on every new |
| 104 |
> Gentoo install. |
| 105 |
> |
| 106 |
> |
| 107 |
> Well if instructions for setting up git sync and |
| 108 |
> associated pros/cons |
| 109 |
> are not documented anywhere then I won't advise anyone |
| 110 |
> to use it. |
| 111 |
> |
| 112 |
> |
| 113 |
> I've attempted to configure it for myself, and this is what |
| 114 |
> it does: |
| 115 |
> |
| 116 |
> * Using keys from /usr/share/openpgp-keys/gentoo-release.asc |
| 117 |
> * Refreshing keys from keyserver ... |
| 118 |
> [ ok ] |
| 119 |
> * No valid signature found: unable to verify signature |
| 120 |
> (missing key?) |
| 121 |
> |
| 122 |
> |
| 123 |
> |
| 124 |
> Please report a bug and attach your configuration along with keyring |
| 125 |
> version. |
| 126 |
> |
| 127 |
> |
| 128 |
> It works after upgrading to openpgp-keys-gentoo-release-20180706 from |
| 129 |
> openpgp-keys-gentoo-release-20180323. |
| 130 |
> |
| 131 |
> |
| 132 |
> Does Portage not call attention to critical updates? |
| 133 |
|
| 134 |
No, but that might be a nice feature. We'd have to introduce some kind |
| 135 |
of standard mechanism via PMS or a GLEP. |
| 136 |
|
| 137 |
> It used to make a special statement for a new stable Portage and |
| 138 |
> strongly recommended that it be emerged first. It should probably do the |
| 139 |
> same for openpgp-keys-gentoo-release. |
| 140 |
|
| 141 |
Sure, but it this case we have a chicken-and-egg problem, because I |
| 142 |
needed the latest openpgp-keys-gentoo-release installed but in order to |
| 143 |
do that I had to sync, but then verification failed because I didn't |
| 144 |
have the latest openpgp-keys-gentoo-release. |
| 145 |
-- |
| 146 |
Thanks, |
| 147 |
Zac |
Archives